Motion control of mobile robots with Particle Filter Model Predictive Equilibrium Point Control

This paper describes an optimal motion control for mobile robots with differential drive. We propose a novel method called Particle Filter Model Predictive Equilibrium Point Control (PF-MPEPC) that implements robot navigation with obstacle avoidance in a structured environment. This control evaluates the near future behaviour continuously to generate an optimal trajectory that is safe and smooth. The Model Predictive Control approach minimizes a defined cost function by solving a non-linear optimization problem. It evaluates robot dynamics and control constraints. Our method uses the Equilibrium Point Control approach. It searches for an equilibrium point near the robot in order to satisfy the navigation requirements. Moreover the use of particle filters allows safety to be improved. This method is validated in a real case scenario with simple obstacles. The obtained results demonstrate that the robot executes safe, smooth and comfortable trajectories.